Polytechnic Institute, Brooklyn, N. Y.
D
URING a study of the correlation between vapor-liquid equilibria and partial heats of solution, the vapor-liquid equilibria for the system water-acetic acid were determined at pressures of 760, 500, 250, and 125 mm. of mercury.
The method adopted for experimentally determining the vapor- liquid equilibria has been described (7, 9, 10). The temperature and pressure are directly observed, whereas the composition of
Figure 1. Arrangement o f Apparatus for Subatmospheric Pressure For use a t a tm o s p h e ric pressu re th e a p p aratu s t o th e rig h t o f s to p co c k E a n d o f th s
reservoir ■was rep la ced w ith a s im p le in d in e d -t u h e m a n om eter.
the liquid and vapor states in equilibrium are determined by suitable analysis of samples of the liquid withdrawn from the still and distillate trap, respectively. The correlation of these data have already been discussed (.11).
The details of the equilibrium still were given in another paper (9); the heating arrangement is shown in Figure 3 of that paper. The coil was wrapped around the larger of the reflux heating tubes between layers of asbestos cloth and was rated at 500 watts (considerably more than needed). Power input was controlled by a Variac transformer.
The arrangement of the apparatus is shown in Figure 1, as prepared for use at subatmos
pheric pressures. For work at atmospheric pressure, the constant-pressure device was re
placed by a small mercury manometer having a slope of about 1 to 10 with the horizontal, and open to the atmosphere. The pressure on the system was then maintained at 760 mm. by- adding air to or venting air from the reservoir through cock E. The column of the still was lagged with many layers of cloth. Cock A served to expel noncondensable gases; cocks B and C allowed liquid and condensate samples to be withdrawn; and the three-way cock D, venting to the atmosphere, served to isolate the still from the reservoir while samples were being withdrawn. The 12-liter flask, used as a reser
voir, maintained steady conditions by absorbing fluctuations of pressure. Both absorption tubes contained soda lime and calcium chloride to
Weighl V. H20
Figure 2. Phase Equilibria of Acetic Acid- Water System
U pper lin e a t 'e a c h pressure is f o r vapor a n d low er lin e is fo r liq u id , b o th p lo tte d a ga in st tem p era tu re.
Ta b l e I. Va p o r- Li q u i d Eq u i l i b r i a Ob s e r v a t i o n s (i n We i g h t Pe r Ce n t Wa t e r) .---P - 760 M m .--- » ,---P mm 500 M m .—» ---P - 250 M m .--- ---P - 125 Mm.— »
T ° C. % * % V T ° C. % * % V T° C. % x % V T ° C. % * % V
118.4 0 0 105.2 0 0 8 5.5 0 0 67.8 0 0
117.6 0 .4 6 1.15 99.55 5 .0 9 .1 80.42 4 .6 8 .3 64.25 3 .9 7 .1
109.1 8 .9 15.6 94.28 14.0 23 .3 77.33 10.5 17.5 61.53 10.7 17.0
105.6 17.3 28.1 9 2.5 4 2 2 .8 34 .8 75.86 1 6.5 2 5 .5 59.83 2 1.4 29.8 103.0 32.85 4 7.6 91.43 3 1.6 45 .4 75.12 2 0 .8 3 0.7 58.61 3 5.6 44.2 101.6 47.4 6 2 .4 90.74 4 0.7 55 .0 73.76 3 4.2 44.6 58.10 4 5.7 53.8 101.2 55.75 6 8 .5 89.94 5 1.8 65 .4 73.09 4 5.0 55.0 57.58 60.1 66.2 100.9 6 3 .6 7 4 .5 89.53 6 3 .4 74 .0 72.51 6 1.0 6 8.2 57.28 7 1.3 7 5.6 100.69 7 4.3 81.7 89.20 7 4.9 82 .2 72.23 7 0 .8 7 6.1 56.92 8 2.2 85.4 100.37 84.0 88.6 88.98 86.6 90 .8 71.87 83.9 87.5 56.68 9 0 .2 9 2 .5 100.12 9 5.4 96.8 88.83 9 5.5 97. 0 71.63 9 5.7 9 3.8 56.57 9 4.8 96.3
100.00 100 100 8 8.7 100 100 7 1.6 100 100 56.4 100 100
liquid; y ■* vapor.
Ta b l eII. Sm o o t h e d Da t ao f Va p o r- Li q u i d Eq u i l i b r i a f o r Sy s t e m Wa t e r- Ac e t i c Ac id0
/— Liquid— » 760 mm.
v » p u r
500 mm. 250 mm. 125 mm.
X I i * ; T ° C. Vi V>[ T ° C. y\ u >1 T° C. in w i T° C. Vi w i
5 1.5* 115.7 10.1 3.2* 103.0 10.1 3.2* •83.3 10.2* 3.3o 6 5.9 10.2» 3.3*
10 3 .2 114.3 18.1 6 .2 101.8 18.0 6.1» 8 2 .3 17.7 6.0» 65.0 17.6 6 .0
15 5 .0 111.9 2 5.5 9 .3 99.6 2 5.0 9 .1 80.1 2 4.5 8 .9 6 3.6 24.3 8 .8
20 7 .0 110.3 3 2.7 12.6 9 8.0 3 2.3 12.4 7 8.8 31.4 12.0 62.7 3 0 .8 11.7
30 11.3 107.8 4 4.7 19.4 9 5.5 4 4.2 19.1 7 7.1 4 3 .4 18.6 61.4 4 2.1 17.8
40 16.6 105.8 55.5 27.1 9 3 .8 54.9 2 6.6 75.9 5 3.5 25.6 6 0.5 5 2.1 24.4
50 2 2 .9 104.4 6 4.9 3 5.6 9 2 .5 6 4.2 3 4.9 7 4.9 6 2.3 3 3.0 59.7 60.7 3 1 .5 60 3 0.9 103.2 7 3.7 45.4 91.5 7 2.8 44.3 7 4 .0 7 0.3 4 1 .4 5 9.0 6 8 .8 39.7
70 4 1.0 102.2 81.3 56.4 90.7 80.7 5 5.3 73.3 78.1 51.3 58.3 76.7 4 9.4
80 54.3 101.3 8 7.5 6 7.6 8 9.9 8 7.3 67.2 7 2.7 85.0 6 2.9 57.7 8 4 .0 6 1.2 85 6 2 .8 101.1 9 0.4 7 3.9 8 9.6 9 0.2 7 3.6 7 2 .4 88.5 6 8.8 5 7.5 8 7.8 6 8.4 90 7 3 .0 100.6 9 3.4 80.9 89.2 93.3 8 0.7 7 2.2 9 2.3 7 8.2 57.2 9 1.8 7 7.0 95 * 8 5 .0 100.3* 9 6.6 8 9 .3 89.0* 9 6.7 8 9.6 71.8* 9 6.3 8 8 .6 5 6 .8» 9 6.0 8 7.9 9 7.5 91.9 1 00.1. 98.2* 9 4.3 88.8* 9 8.3 94.5 71.6* 98.2* 94.3 56.6» 9 8.2 94.0 99 96.7* 100.0* 9 9 .2 . 9 7 .6i 88. 7ï 99.3» 97.8* 7 1 .6i 99.3» 97. 7t 5 6 .5o 99.2* 97.6*
fl x *•mole % in liquid ; y - mole % in vapor ; w - * weight % in liquid; u>' — weight % in vapor; sub-jc r ip t 1 ■» water.
The vapor-liquid equilibria and boiling points were determined for the binary system water-acetic acid at the constant pressures o f 760, 500, 250, and 125 m m . of mercury by the method previously described (9). The experimental data are presented in graphs and tables, including one for smoothed data obtained from the graphs.
A comparison with other data in the literature is given for boiling points and vapor compositions at atmospheric and subatmospheric pressures.
protect the mercury from acid or water vapor. Tw o mercury traps prevented any mercury from being lost if sudden surges took place; the water trap prevented any water from the aspirator from being sucked into the system if sudden changes in water pressure took place. Three-way cock E was used for rapidly evacuating the system to approximately the correct pressure before suction was permitted by way of the pressure regulator, which maintained the pressure at any desired value.
Cock F of the constant-pressure regulator (4) enables the cor
rect amount of mercury to be added to thfe manometer to ob
tain the arbitrarily chosen pressure.
Ch a r g i n g St i l l. The still was charged with about 200 cc.
of pure acetic acid, and the boiling point was determined at each pressure. By adding the correct amount of distilled water and draining the corresponding amount of solution after each run, about ten evenly spaced values of weight per cent water for the liquid composition were obtained for each pressure.
Co r r e c t Pr e s s u r e. The system was closed off from the atmosphere by turning cock D so as to connect the still with the reservoir. With cock A opened, the system was evacuated through cock E with the pressure regulator by-passed. When the desired pressure was nearly reached, cock E was turned so that all aspirated gas would have to pass through the regulator which had been pre- _________________ viously set at the desired pres
sure. With the system at the cor
rect pressure, the liquid was brought to a steady boil by reg
ulating the heat input by means of the transformer. When all non- condensable gases appeared to have been bled through valve A, it was closed.
Eq u i l i b r i u m. Boiling was al
lowed to proceed until equilibrium was reached at the desired pressure, as indicated by a constant reading o f the thermometer. The average time to come to a steady state was usually about 20 minutes, although at least 40 minutes were allowed. The temperature of the system in equilibrium was taken when the manometer reading drifted to the exact value at which the pressure was previously chosen.
Two check readings, 10 minutes apart, were used as a criterion for concluding the run.
After equilibrium conditions were verified by the check reading of the temperature, heating was stopped;
the still (but not the reservoir) was opened to the atmosphere through cockD.
1063 Wi t h d r a w a l o f Sa m p l e s. Samples of the liquid in the
still and in the reservoir were withdrawn into previously weighed flasks, each containing 25 cc. of approximately 0.2 N barium hydroxide and 2 drops of phenolphthalein, just to re
move the pink coloration. The flasks were immediately stop
pered and weighed. For the first set of runs, two samples were withdrawn from both the still and condensate trap as a check on the method, which was found to be more accurate and con
venient than any other attempted.
The titrating buret was so arranged that solution could be added directly without any contact with the air. The solu
tion was standardized against certified potassium acid tartrate obtained from the National Bureau of Stand
ards. From the titers, the compositions of the vapor and liquid in equilibrium at the given pressure and measured temperatures were obtained.
-S' C A L IB R A T IO N S AND S M O O TH IN G
O F DA TA The 0.1° C. thermometer for subatmospheric runs was cali
brated by the National Bureau of Standards for total immersion; a stem correction
F ig u r e 3. Vapor Com position v s . Liquid Com posi
t io n fo r A c e tic Acid-Water System
0 10
was always applied in the usual manner. The 1° C. ther
mometer for atmospheric runs was calibrated against the 0.1° thermometer for total immersion, and a stem correction was applied. A magnification glass was employed to read the thermometers so that tenths and hun
dredths were estimated on 1° and 0.1° thermometers, respec
tively.
The temperature and pressure measurements were calibrated by taking the boiling points of distilled water at several pressures.
The pressure reading was corrected for vapor in the manometer by comparison with the barometer reading. The agreement between the observed and true boiling points was excellent;
between the range 55° to 85° C., the deviation was less than 0.2°. With increasing temperatures the boiling point read too high and was in error by 0.7° at 100° C.
Titration showed the reagent-grade acetic acid to contain 99.8% acid by weight; the bulk of the remaining constituents were assumed to be water as the total amount of other con
stituents according to analysis was less than 0.01%
by weight. The water used in the experiments was freshly pre
pared distilled water.
The first set of runs at 760 mm.
(Table I) employed the 1 ° C. ther
mometer; a second set of runs on dilute acid solutions used the 0.1° thermometer and was performed after subatmospheric data were obtained. In the first set two samples
^ each of the liquid and vapor compositions were analyzed to check the method of analysis. In most cases the check was within 0.1% weight concentra
tion and never exceeded 0.2%. The accuracy of the titrations tend to increase as the concentration of acid is diminished, mainly because the size of sample necessarily increases for the same volume of standard solution and also because the loss by evaporation of acid is reduced.
A graph of phase equilibria against temperature is given in Figure 2, and of vapor against liquid composition, in Fig
ure 3. From these graphs the data were smoothed as in Table II.
C O M PA R ISO N W ITH O T H E R DATA
To t a l Pr e s s u r e a n d Bo i l i n g Po i n t s. For the comparison of boiling points as a function of pressure and composition, the log of total pressure was plotted in Figure 4 against the log of vapor pressure at constant liquid composition, as previously described (8). Included are data from two other sources. The data of Keyes (6) show deviations since the experimenter was essentially concerned with vapor composition rather than temperature measurement; the data of Kahlbaum and Konowalow (5) are in better agreement. The latter investi
gators used the static method of determining total pressure at a given temperature which gives larger errors at decreased pressures.
Va p o r- Li q u i d Eq u i l i b r i a a t At m o s p h e r i c Pr e s s u r e. A large plot was made of (y — x) against x (Figure 5) to compare the experimental values of this investigation with those from available literature (7, 2, 3, 7, 12-16). The difference plot emphasizes small variations. Wherever possible, the original observations of the investigator were chosen rather than smoothed data.
Another article (11) has discussed the correlation of these data by means of heats of solution and by means of several new methods of plotting developed for handling these p-t-x-y data.
Tem perature *C.
Vapor Pressure W ater mm.Hg
Figure 4. T ota l V apor Pressure o f A cetic A cid -W a ter M ixtures vs. Vapor Pressure o f W ater o n L og a rith m ic C oordinates Outa o f K a h lb a u m a n d K on ow a low a n d o f Keyea co m p a re d at 25« 50« and
75 % w eig h t water.
A C K N O W LE D G M E N T
Thanks are due A. J. Marsh and B. H. Krehbiel for supplying laboratory facilities and assistance at the McKinley Technical High School, Wash
ington, D . C., where the experimental work was done.
UTERATUR EXCITED
(1) Bergstrom, quoted by Hausbrand In “ Principle«
and Practice of Industrial Distillation” , 4th ed„ p. 238 (1925).
(2) Blacher, Ibid., pp. 238—42.
(3 ) Cornell, L. W., and Montonna, R . E., In d. En o. C h e m ., 2 5, 1331 (1 9 3 3 ).
(4) Gilmont, R., and Othmer, D. F., I n d . E n o . Ch e m., An a l. Ed., 15, 641 (1 9 4 3 ).
(5) Kahlbaum and Konowalow, International Critical Tables, Vol. I ll, p. 306 (1928).
(6) Keyes, D. B., I n d . E n o . C h e m ., 2 5, 669 (1933).
(7 ) Othmer, D. F., Ibid., 20, 743-6 (1928).
(8) Ibid., 3 2, 841 (1 9 4 0 ).
(9 ) Ibid., 35, 614 (1 9 4 3 ).
(10) Othmer, D. F., I n d . E n o . C h e m ., A n a l . E d ., 4, 232 (1932).
(11) Othmer, D. F., and Gilmont, R „ I n d . E n o . C h e m ., 36, 858 (1 9 4 4 ).
-(12) Pascal, Dupuy, Ero, and Garnier, Bull. §o*.
chim., 29, 9 (1921).
(13) Povarnin, G., and Markov, V., International Critical Tables, Vol. I ll, p. 310 (1928).
(1 4) Rayleigh, Phil. Mag.; [6] 4 , 5 2 1 -3 7 (1 9 0 2 ).
(15) Sorrel, E., Compt. rend., 122, 946 (1896).
(16) York, R., Jr., and Holmes, R . C.. In d . E n o . Ch e m.. 3 4, 345 (1 9 4 2 ).
Figure 5. D ifference in Vapor and L iquid C om p osition s (in M ole Per C ent) o f A cetic A cid—W ater System r*.
M ole Per Cent W ater in Liquid o O th m e r
n Y ork 5 H olm es v C o r n e l ia M o nton no a B e rg s tro m
• R a y le ig h
■ B lo c h e r
a P ovarnin a M o rkov
* P a s c a l, Dupuy, a G arnier e S o rre l